Why not both? Rethinking categorical and continuous approaches to bilingualism.
Why not both? Rethinking categorical and continuous approaches to bilingualism.
复制标题
DOI：
10.1177/13670069211031986
复制
发表时间：
2021-12
期刊：
影响因子：
--
通讯作者：
Byers-Heinlein K
中科院分区：
文献类型：
--
作者：
Kremin LV;Byers-Heinlein K
双语是一个复杂的结构，很难定义和建模。本文提出，双语领域可以借鉴其他心理学领域，通过整合先进的心理测量模型，包括分类和连续性质。这些模型可以将文献中存在的双语和单语群体的广泛使用与最近提出的双语应被视为一个连续变量的建议统一起来。在本文中，我们强调了两个潜在的兴趣模型：因子混合模型和隶属等级模型。这些模式同时允许不同类别的说话者的形成，以及在这些类别中存在的连续变化。我们讨论了如何在双语研究中实施这些模型，包括如何开发这些模型。当使用这两种模型中的任何一种时，研究人员可以对分类信息或连续信息或两者的组合进行分析，这取决于哪一种最适合解决他们的研究问题。将更复杂的模型纳入双语定义可以使双语研究领域受益。为了帮助双语研究的各个子领域集中在合适的模型上，我们鼓励研究人员预先注册他们的模型选择和计划分析，并分享他们的数据和分析脚本。本文独特地提出结合先进的统计心理测量方法来定义和建模双语。在这些更灵活的模型的背景下概念化双语将允许解决各种各样的研究问题。最终，这将有助于推进理论，并导致对双语的更全面、更深入的理解。
Bilingualism is a complex construct, and it can be difficult to define and model. This paper proposes that the field of bilingualism can draw from other fields of psychology, by integrating advanced psychometric models that incorporate both categorical and continuous properties. These models can unify the widespread use of bilingual and monolingual groups that exist in the literature with recent proposals that bilingualism should be viewed as a continuous variable. In the paper, we highlight two models of potential interest: the factor mixture model and the grade-of-membership model. These models simultaneously allow for the formation of different categories of speakers and for continuous variation to exist within these categories. We discuss how these models could be implemented in bilingualism research, including how to develop these models. When using either of the two models, researchers can conduct their analyses on either the categorical or continuous information, or a combination of the two, depending on which is most appropriate to address their research question. The field of bilingualism research could benefit from incorporating more complex models into definitions of bilingualism. To help various subfields of bilingualism research converge on appropriate models, we encourage researchers to pre-register their model selection and planned analyses, as well as to share their data and analysis scripts. The paper uniquely proposes the incorporation of advanced statistical psychometric methods for defining and modeling bilingualism. Conceptualizing bilingualism within the context of these more flexible models will allow a wide variety of research questions to be addressed. Ultimately, this will help to advance theory and lead to a fuller and deeper understanding of bilingualism.
登录
查看更多内容
影响因子：
3.7
作者：
Andreotti A;Minicuci N;Kowal P;Chatterji S
通讯作者：
Chatterji S
影响因子：
2.1
作者：
Baum, Shari;Titone, Debra
通讯作者：
Titone, Debra
影响因子：
1.8
作者：
Baker, W;Trofimovich, P
通讯作者：
Trofimovich, P
影响因子：
2.6
作者：
Bosch, Laura;Sebastian-Galles, Nuria
通讯作者：
Sebastian-Galles, Nuria
影响因子：
4.1
作者：
De Cat, Cecile;Gusnanto, Arief;Serratrice, Ludovica
通讯作者：
Serratrice, Ludovica